Comparing American and British English

different from, than or to?

A superficial analysis of the graph

Wikipedia (I know, not the most reliable, but certainly a very common source) writes:

In both dialects, from is the preposition prescribed for use after the word different: American English is different from British English in several respects. However, different 'than'''''' is also commonly heard in the US, and is often considered standard when followed by a clause ("American English is different than it used to be"), whereas different to is a common alternative in BrE, despite its informality.

(Wikipedia, March 11, 2017.)

Summary of Wikipedia's claim:

  1. different + from is the most common in both
  2. different + than is used frequently in spoken AmE
  3. different + to is preferred over different + than in (spoken?) BrE

Google's N-Gram Viewer gives us a closer look at the usage of written (!) language over time. The following graph compares the frequency of different from to different to and to different than across the two main varieties of English: BrE (eng_gb_2012) and AmE (eng_us_2012) during the last 200 years. As you can see below, the usage preferences for different + preposition seem near identical in the two varieties, but different + than is more frequent in written AmE than in BrE and the opposite is true for different + to:

  1. correct: different + from is the most common in both
  2. written evidence supports this claim: different + than is used frequently in spoken AmE
  3. written evidence supports this claim: different + to is preferred over different + than in (spoken?) BrE

While different + from has been greatly favoured in writing for the past 200 years, there is a slight trend in favour of different + than/to and at the cost of different + from (at least in AmE) that seems to have started in the 1960s.

Obviously, this is just an illustration of how Google's N-Gram Viewer can be used for illustrating (supposed) similarities and differences between different varietes of English (also see the corpus for English fiction). It does not give you immediate access to date for statistical anylisis. Nevertheless, it allows you to quickly compare expressions across time and space and make it immediately visually accessible.

How to do it yourself (micro-step-by-micro-step)

Google Books provides a detailed overview of the various functions of the Ngram Viewer. In a nutshell, this is what I did:

  1. Open another text pogram (e.g. MS Word, notepad, ...) to type in the search terms because this can get really complicated and the searchbox can get too short.
  2. Look up the corpora names on the info site about the Ngram Viewer, section Corpora, and pick corpora from the table (e.g. eng_us_2012 and eng_gb_2012).
  3. Copy-paste the corpara shorthand into your text program, as separate paragraphs, like a list:
        1. eng_us_2012
        2. eng_gb_2012
  4. Now add colons before the corpus shorthands and commas after:
        1. :eng_us_2012,
        2. :eng_gb_2012,
    1. Multiply by as many search terms as you want to compare, in this case 3 (different from/to/than) and remove the last comma. We now have six entries (2 corpora * 3 search terms):
        1. :eng_us_2012,
        2. :eng_gb_2012,
        3. :eng_us_2012,
        4. :eng_gb_2012,
        5. :eng_us_2012,
        6. :eng_gb_2012
  5. Now insert the (3) search terms in twos, once for each corpus, just before the colon of each corpus, but add no spaces (but the search term may contain spaces):
      1. different from:eng_us_2012,
      2. different from:eng_gb_2012,
      3. different than:eng_us_2012,
      4. different than:eng_gb_2012,
      5. different to:eng_us_2012,
      6. different to:eng_gb_2012
  6. Delete the paragraph markers and create one line of serch terms, without spaces around the commas or the colons.
  7. Paste this into the search box at the top of the NGram Viewer and press enter.